machine learning

Apr
19
Distributed Training

Distributed Training

This post is about distributed training with pytorch on sagemaker.
Apr
19
SageMaker Inference

SageMaker Inference

Sagemaker inference provides 4 different types of inference: batch, serverless, real-time, and asynchronous. This is some basic pytorch code: import
Aug
03
CLIP Model

CLIP Model

Creating a customer script to run CLIP on sagemaker. # clip/code/inference.py import base64 from io import BytesIO from
1 min read